By carrying out density functional theory (DFT) calculations for 60 polyvinyls, polyethylenes and polymethacrylates repeating units at the B3LYP/6-31G(d) level, two quantum chemical descriptors, the molecular traceless quadrupole moment Θ, the molecular average hexadecapole moment Φ, are obtained and used to predict the glass transition temperature ( T g). A physically meaningful quantitative structure–property relationship (QSPR) model having correlation coefficient 0.953 for the training set and 0.952 for the test set is generated using multiple linear stepwise regression analysis. It suggests that the model obtained here can predict the T g values of polymers and provide theoretical guidance for polymeric molecular designs. Compared with the existing QSPR models, the proposed model with only two multipole moment descriptors is most simple. This paper encourages the further application of multipole moments to the QSPR studies for polymers.